Hello,
IMO the OB 'magic' is not primarly due to front wall reflection, but the higher total direct to reflection ratio at the listening position compared to a monopole source expecially in the bass-midrange below about 1kHz. This allows better reproduction of the original recorded 'wave' at the listening position because less of room masking by reflections.
- Elias
IMO the OB 'magic' is not primarly due to front wall reflection, but the higher total direct to reflection ratio at the listening position compared to a monopole source expecially in the bass-midrange below about 1kHz. This allows better reproduction of the original recorded 'wave' at the listening position because less of room masking by reflections.
- Elias
Last edited:
The Soren Bech tests simulated a speaker in a domestic room while varying the level of every reflection in turn. He found the floor bounce had a threshold of -6.5dB relative to the direct sound, but a typical level of -3.5, i.e. it was audible. His tests were to detect timbral changes in pink noise and speech. I'm not sure what the +5 to +10 figures refer to.
In general the detection level drops (things are more audible) as time goes on, there is heavy masking for the first several msec, but at the same time later reflections will have gone a greater distance (level drop with distance) and will have traversed several boundaries (boundary attenuations). So there is a catch 22, later reflections might have potential to be more audible but typically they are rapidly dropping in level. In the end, Bech found the floor bounce was singularly likely to be audible, at 3dB above the threshold of detection.
David S.
You shouldn't read what Toole had to say about dipoles, then. When he did blind tests with the ESL63 dipole grouped with other non-dipole speakers the ESL63 did poorly, especially in mono. It was downgraded for a lack of spaciousness due to its higher directivity. In the stereo tests it was still downgraded but not quite as badly.
I suspect that higher directivity CD horn systems would suffer similarly in those tests.
Come back Bose 901, all is forgiven.
David S.
The Bech study is done with pink noise with a level at the of listeners of 66dB. Speech was 50dB (fast time weighting). These are fairly low levels. I'm not sure if there are effects due to level.
David S.
Hello,
His ranking is based on listener preference. People can have different preferences.
In my experience the 'spaciousness' is a high freq phenomena in the home listening environment, somwhere above 1-2kHz. ESL63 is quite narrow directivity in the treble area, that is in comparison with the other speakers used by him which propable used dome tweeters with much wider dispersion, and thus the result does not directly suggest that dipole did worse (in listener preference) but that the high freq range directivity should be wide (in order to gain points in listener preference) in home listening environment.
- Elias
His ranking is based on listener preference. People can have different preferences.
In my experience the 'spaciousness' is a high freq phenomena in the home listening environment, somwhere above 1-2kHz. ESL63 is quite narrow directivity in the treble area, that is in comparison with the other speakers used by him which propable used dome tweeters with much wider dispersion, and thus the result does not directly suggest that dipole did worse (in listener preference) but that the high freq range directivity should be wide (in order to gain points in listener preference) in home listening environment.
- Elias
Toole makes an interesting notion about the dipole:
"According to these data, in all but one category, loudspeaker BB
received lower (only slightly but consistently lower) spatial ratings than the other
two products. Only in the category “abnormal spatial effects” was the rating
higher, and for this category, that is a problem (most often sounds were criticized
for being inappropriately close to the listener and occasionally inside the head;
see “In-Head Localization”)."
"According to these data, in all but one category, loudspeaker BB
received lower (only slightly but consistently lower) spatial ratings than the other
two products. Only in the category “abnormal spatial effects” was the rating
higher, and for this category, that is a problem (most often sounds were criticized
for being inappropriately close to the listener and occasionally inside the head;
see “In-Head Localization”)."
... and:
"Interestingly, a demonstration
of four-loudspeaker Ambisonic recordings played
in an anechoic chamber yielded an auditory impression
that was almost totally within the head. This was a great
disappointment to the gathered enthusiasts, all of whom
anticipated an approximation of perfection. It suggested
that, psychoacoustically, something fundamentally important
was not being captured or communicated to the ears.
An identical setup in a normally refl ective room sounded
far more realistic, even though the room refl ections were a
substantial corruption of the encoded sounds arriving at
the ears."
"Interestingly, a demonstration
of four-loudspeaker Ambisonic recordings played
in an anechoic chamber yielded an auditory impression
that was almost totally within the head. This was a great
disappointment to the gathered enthusiasts, all of whom
anticipated an approximation of perfection. It suggested
that, psychoacoustically, something fundamentally important
was not being captured or communicated to the ears.
An identical setup in a normally refl ective room sounded
far more realistic, even though the room refl ections were a
substantial corruption of the encoded sounds arriving at
the ears."
This is an indication to me that there is a range of dipole designs that will each provide differing perceptions. Flat panel dipoles such as electrostats are not going to operate as dipoles very high in frequency. Long ribbon dipoles with narrow baffles would do much better, but suffer from vertical response issues with even small changes in height. Multi-way dynamic driver dipoles with incrementally smaller baffle widths/drivers are going to operate as true dipoles to higher frequencies.
My old Apogee Centaurus Ribbon hybrids, dipole ribbons from 200Hz up were my old favorites largely due to the spaciousness and depth when placed well away from side walls (>10ft. in my old house) and at least 3ft from the front wall. They weren't pure dipole due the ribbon being near an edge of a modestly wide baffle and the long ribbon definitely posed problems in setup.
I would offer that multi-way narrow or tapered baffle dipoles are the best candidate for comparison due to dipole behavior from both baffle width and incrementally smaller driver diameter resulting in better DI to higher frequencies. Research limited to dipoles such as the ESL63 are not nearly as relevant today, at least IMO. Better dipoles exist today, just as do better waveguides, etc.
Dave
OB magic, eh?
Have you ever listened to an OB speaker with and without the front wall? I have. It's very interesting what those reflections do - or don't do. It might surprise you.
Yes, thanks Marcus for the good quote.
Have you ever listened to an OB speaker with and without the front wall? I have. It's very interesting what those reflections do - or don't do. It might surprise you.
Yes, thanks Marcus for the good quote.
Not sure about this. I would think that, as a class, dipoles have more uniformity than most other configurations. Electrostats are clearly dipoles for all frequencies (equal front and rear radiation plus sideways nulls). At higher frequencies their front/rear radiation may narrow and break into multiple lobes, but that would only make matters worse in the Toole tests.
Based on his tests the general conclusions are that somewhat delayed lateral reflections add a pleasing sense of spaciousness and even work in concert with recorded ambience. Dipoles can be a poor choice because they accentuate back wall bounces, which add colorations not easily seperated from the direct sound, and downplay lateral reflections that are helpful.
Operating as a true dipole over a wide range isn't necessarily a virtue. As audiophiles we need to learn to separate idealistic models (omni directional, pulsating sphere, linear phase, high amounts of low order distortion, high efficiency, constant directivity, flat sound power, etc., etc.) from the real performance attributes that contribute to realism.
David S.
In my experiences, electrostatics are wholly separate from dynamic (and even planar/ribbon) dipoles. The real world listening experience can be night and day when it comes to directionality. I would never mistake a pair of Acoustats for dynamic dipoles, or even a pair of Maggies.
Also, once dipoles are brought out into the room, as they should be, the argument about front wall reflections being accentuated starts becoming a little overdone -- especially in a room deemed a "listening room". It's not hard to tune "to taste".
That they can be a poor choice does not mean that they are always a poor choice. Monopoles can also be a poor choice or the lesser choice vs. other configurations. Room and placement conditions cannot be neglected when debating choice and characteristics.
I'm not sure that as a class uniformity would be higher. I am thinking more in terms of the full range figure-of-eight dipole response with fewer lobes. This is not limited to the rear radiation and back wall reflections. Wider dipoles are going to start to lobe earlier, so the off-axis response, not just the side 90-degree point, will become more erratic, so-to-speak, and has an equal impact on the front, direct response. For a multi-way dipole system, driver selection and crossover are significant factors. The transition from full figure-of-eight to a more directional response is going to create a transition in the off-axis polar response, so different dipole designs will have different sidewall reflected spectra as well. But this part is no different than other types, such as 2-way monopoles vs. multi-way monopoles and off-axis flares. Horizontal polar response lobes in a wide dipole can be worse than monopole polar response, at least when comparing the front, direct response.
Smoothness of the polar response is my concern and my focus. As far as back wall reflections, this is only a concern of distance and degree, since even monopoles exhibit the same symptom starting in the baffle step area and below. I have never found a monopole to have as good a sense of spaciousness as my dipole ribbons did when the latter were placed appropriately. The key is placement. Dipoles have different requirements, but as long as those placement requirements are met, they are IME superior in certain areas.
Narrow dipoles are going to radiate more to the rear, yes, but narrow dipoles with fewer lobes will also have a smoother polar response that should provide more more uniform response of any reflected sound. I suggest that as far as dipoles specifically are concerned, narrow is better than wide. It is possible to create a multi-way dipole that does not exhibit lobes in the horizontal polar response prior to the point where the high frequency driver directionality takes over. I suggest that this would improve the results in Toole's tests were they to be used. I also think it's better to compare the perceived response of each configuration when placed in its optimum conditions if optimum performance is the criteria.
Dave
Edit: An ideal dipole has no response at 90 degrees, but this is not the angle at which we are subjected to reflected responses from side walls. These come from some point that is at an angle somewhat in front of the system, dependent upon the placement of system and listener, the area at which lobes become an issue. The dipole null helps to reduce system power response, but does not eliminate all side wall reflections.
Last edited:
Glad you asked. 😉
I've posted this before, but it bears repeating.
With no wall behind an open baffle speaker (I mean none at all, but the other walls in place) the effect is remarkable. The depth of the image can be several meters behind the speakers. Imaging is good, stable and there is a "That's Real!" feeling to the sound.
Replace the wall and depth is almost all gone. The live feel is greatly diminished and distortion seems to rise.
However, there is much more energy in the bass and mid bass - a much fuller sound. The effect is similar with monopole boxes, but not nearly as pronounced.
The wall behind the OBs does both good and bad. It can kill good imaging and depth, but does return a lot of useful energy, allowing fuller sound. It seems to me that heavy diffusion behind the speaker enjoys the best of both worlds. No coherent reflections to muddle the image and "liveness" but it can retain the lower frequency energy needed for a good tonal balance and impact.
That would be the front wall, the wall you're looking at from the listening position?
So is it a very early reflection coming from the same direction as the direct sound that prevents perception of depth?
Other report increased realism in depth perception by adding reflections instead of removing them.
Do you think the contradiction might be result of how long delayed the reflection is from the front wall? That is, noticeable but fainter sound (than direct) gives the impression of depth?
Have you looked at J. Usher's thesis? He's done a lot of work on decorrelation for purpose of creating better surround sound from 2 channel original signal.
Chapters 2 and 3 are about spacial imaging. I think you might find it interesting.
http://www.jar-lab.com/papers/PhD/Usher_PhD.pdf
Last edited:
- Status
- Not open for further replies.